Will the ToE AI be less dumb??

[MistWeaver]

I think it's clear that the AI will never be a challenging pseudo-human player.

Once again: try GalCiv2 Dark Avatar with latest update (1.6 beta 6, at this moment) with max CPU using option turned on. Its close to that.

[krs]

On of the reasons AI looks so stupid is poor quality of scripting functions for map editor. (I strongly suspect they left that part to the cleaning lady.)

Just an example:

On the heavan campaign mission 3? (I think), you are under siege by Inferno heroes. AI priorities are clear: Take the heaven castle, NO MATER WHAT! but If you have a stronger hero standing in AI's path, the AI instead of attacking everything in its path, will just go wandering through the map until the "coast is clear". If you keep the hero in the castle, AI has no problems jumping at your neck.

To bad for the map makers .

[Alamar]

The resources that the enemy makes you waste are "creatures in your army" which is a limited resource.

In a case where the AIs are basically allied against you turtling could become a winning tactic in the right circumstances.

Wrong. you don't need to attack him until his ally is taken care of. Just take his mines, that way he won't be able to buy enough creatures unless he's cheating. And even if he can, your control of adv map creature dwellings and high income means you'll have more eventually.

If he only turtles and doesn't attack you can leave him last.

You're wrong on a couple of points:

1. On Heroic difficulty the AI doesn't need mines in order to totally buy out their creatures and have a lot of resources left over. The AI gets enough resources from a fully built castle to easily buy out its weekly creatures. Because of reduced resource costs you're not really depriving them of much assuming that it at least takes you a while to break out of your home territories ....

2. Bypassing an AI faction that decides to turtle doesn't buy you much at all. When you are no longer "close" the AI can afford to use guerrilla tactics to weaken you economically [I.E. come out of its shell] If you attack an allied faction then there's nothing that keeps them from also turtling so you're going to have to deal with the issue directly sooner or later.

3. If the AIs control more castles than you then they are producing more creatures per week than you are likely to produce. Your best alternative is to go on the offensive and start chipping away at the turtle shell and see what happens.

However I think that this is getting away from the point because I don't think the HoMM5 AI turtles properly [it rarely comes out of its shell] and it makes for boring game play.

Note: I'm not saying that turtling is always a winner ... it's NOT. I'm just saying that there are times where going on the defensive may be excuseable from certain limited perspectives.

--------------------------------------------------------------------------

I think it's clear that the AI will never be a challenging pseudo-human player.

Once again: try GalCiv2 Dark Avatar with latest update (1.6 beta 6, at this moment) with max CPU using option turned on. Its close to that.

The GalCiv AIs have 2 unfair advantages over HoMM:

1. Those games [IIRC] are RTSs which means the AI may not have to be smart but can use its micromanaging and speed advantage to make up a lot of the difference.

2. The GalCiv devs actually care about their AI [unlike UBI / Nival]

[okrane]

turtling + really long turns = boring and a waste of my time

[NurDas]

1. Those games [IIRC] are RTSs which means the AI may not have to be smart but can use its micromanaging and speed advantage to make up a lot of the difference.

I am pretty certain that Galactic Civilization 2 is a turn based strategy. Or am I not getting something

[Kristo]

You're correct. The GalCiv series is a "4x in space" TBS. The reason their AI is so good is that it's been in development since GalCiv1 for OS/2 back in 1993 or so. That, and their lead developer is also the chief AI coder.

[Caradoc]

Of course it is possible to have a challenging AI that plays by human rules. Deep Blue proves that. But the question is how much the developer of a $50 game can afford to invest in the AI.

[ThunderTitan]

Of course it is possible to have a challenging AI that plays by human rules. Deep Blue proves that. But the question is how much the developer of a $50 game can afford to invest in the AI.

Plus, coming to your house after each game is a bit unfeasible.

[MistWeaver]

But the question is how much the developer of a $50 game can afford to invest in the AI.

Enough to make it that good, while supa-3d-awesome graphics(which actualy sucks) is not at first place in the budget

Plus, coming to your house after each game is a bit unfeasible.

[Caradoc]

But the question is how much the developer of a $50 game can afford to invest in the AI.

Enough to make it that good, while supa-3d-awesome graphics(which actualy sucks) is not at first place in the budget

Plus, coming to your house after each game is a bit unfeasible.

Personally, I'd be happy to trade off some of the graphics splendor for a better AI. But the question is would the resulting product be viable? Lack of the cool graphics would impact estimated sales, which would reduce the development budget and there go the resources for improving the AI. From a marketing standpoint, the right thing is an AI that is 'good enough' for most players. Maybe not for you and me, but for most of the potential customers.

An alternative is a $100 product, generating more profits on lower sales. Another is to massively reduce distribution costs by selling on line. But neither of these approaches has proven to work with a niche product like Heroes.

[PhoenixReborn]

AI, ToTE. One thing that pops to mind is how will it handle creature diversity?

It obviously isn't going to think about having a small stack of paladins for cleaning, and a large stack of (champions) for killing, since it already doesn't, but will it jumble alternates on the same tier together, or will it have a nice spread? Again a basic question, that we don't know the answer to until it comes out, but hopefully Nival are thinking about this stuff.

[ThunderTitan]

Lack of the cool graphics would impact estimated sales,

That's why the Wii is doing so bad and the PS3 so bad. And the graphics aren't that cool anyway.

[Orfinn]

Barricading yourself in your town with your main hero, buy out the creatures each week and wait for the opponent to attack you in your town. It's a denying strategy more than a winnig strategy either in the hopes to stalemate or even win a hero battle you'd otherwise lose.
The main thing is that it works only when there is only the "destroy all opponents" VC, which is impractical in my experience in an online or hotseat game against human opponent(s).

Bottom line is that the AI is showing its quality only on maps with other VSs because passive play and turtling won't help winning, while on the other hand other VSs COULD be fulfilled cheating (you can't really control whether the AI really HAD the force to take that town or really DID visit enough monoliths to dig out the tear, do you) or never at all due to the AI not reacting and simply turtling.

Thanks, its clear for me now. Well I was such an turtle player in H3 for a while back, now in any Homm game I put a weaker hero in town with alot of troops but having a stronger main hero with a decent amount of troops with him/her to cause trouble on the map.
But ofcourse if I get cornered its just natural for me to turtle and strenghten my towns defences with whatever I can get of help.

[Alamar]

1. Those games [IIRC] are RTSs which means the AI may not have to be smart but can use its micromanaging and speed advantage to make up a lot of the difference.

I am pretty certain that Galactic Civilization 2 is a turn based strategy. Or am I not getting something

Then I stand corrected.

This still leaves the GalCiv AI with unfair advantages:

1. Their game designers & programmers are capable of designing a game and creating an AI that can play that game.

2. I suspect that there are fewer "bean counters" forcing an unfinished game out of the door.

3. GalCiv developers actually give a darn.

I am not convinced that HoMM has any of these things going for it.

-----------------------------------------------------------------------------

As far as handling various creatures [with the new system] I don't suspect that it will be able to do more than the most primitive / basic imaginable handling .... I.E. the AI will probably do the absolute minimum that it has to in order to make the game run ....

[Fortuna]

I've been waiting on Heroes V (Civ IV helps make me patient). Yesterday I almost bought it, since it's down to $30. Then I remembered how disappointed I was after buying HoMM IV, with its narcoleptic AI, and I figured I should do some research on HoMM V forums first. I guess it's a good thing I did.

What you're all describing here sounds like HoMM IV, or almost as bad, anyway. In other words, it sounds like a waste of money for anyone who doesn't have friends to play head-to-head.

Such a pity. I don't understand how HoMM II and III could have better AIs than IV and V. There's no excuse for it.

[Romanov77]

I've been waiting on Heroes V (Civ IV helps make me patient). Yesterday I almost bought it, since it's down to $30. Then I remembered how disappointed I was after buying HoMM IV, with its narcoleptic AI, and I figured I should do some research on HoMM V forums first. I guess it's a good thing I did.

What you're all describing here sounds like HoMM IV, or almost as bad, anyway. In other words, it sounds like a waste of money for anyone who doesn't have friends to play head-to-head.

Such a pity. I don't understand how HoMM II and III could have better AIs than IV and V. There's no excuse for it.

no, wait.
Its not THAT bad.

AI builds up stuff fine and train its troops and will brutally own you if it sees that his army is stronger than yours.

however sometimes forget to take "freebies" and mines and on some maps (man I need to find what triggers it!!) their turn takes several minutes for no apparent reason.

[MistWeaver]

no, wait.
Its not THAT bad.

AI builds up stuff fine and train its troops and will brutally own you if it sees that his army is stronger than yours.

however sometimes forget to take "freebies" and mines and on some maps (man I need to find what triggers it!!) their turn takes several minutes for no apparent reason.

It is not that bad first two month, after that its actully even worse (HoF version for sure, dont have desire to test 1.5).

[Alamar]

I agree that the AI doesn't seem that bad .... However the AI isn't nearly as good as it should be either.

Basically [on the harder difficulties] the AI has virtually unlimited funds so you can imagine the problems dealing with "bull rushes".

After the first month or two the AI turtles too much. I'm not saying that's the wrong move from certain perspectives but it certainly makes things boring ....

The AI turn duration issue seems to be a hit-or-miss type of bug. You may get it which WILL make the game unplayable or you won't get it much with a properly patched game.

It might not be a terrible idea to wait until the equivalent of an H5 Gold comes out and then wait for that to hit 25$ to 30$ and that would probably be OK.

[Fortuna]

Hrm.

How does the V AI compare to the AI in III?

[MistWeaver]

Hrm.

How does the V AI compare to the AI in III?

Adventure - H3 is better, simply because H5 :
1) Almost never flees from battle, thus losing its best heroes.
2) Becomes passive after ~2 month.
3) Too often ignores mines & chests.
4) Its designed to lose.

Battle - I guess equal. Actualy H5 Battle AI could be better, but sometimes it does very stupid things. Especially in 1.5 as I heard.